[Gluster-users] "Self heal" or sync issues?

Martin Schenker martin.schenker at profitbricks.com
Thu Apr 28 11:54:42 UTC 2011


Hi all!

We're running a 4 node cluster on Gluster 3.1.3 currently.

After a staged server update/reboot only ONE of the 4 servers shows some 
mismatches in the file attributes. It shows that 28 files differ from 
/0x000000000000000000000000/ the "all-in-sync" state. No sync or self 
heal has happened within the last 16h, we checked last night, this 
morning and now.

Even after opening each file with /od -c <file> | head -2/ the 
self-heal/sync process doesn't seem to start.

There are no errors in the logs, how can I check that things ARE 
happening correctly?

/0 root at de-dc1-c1-pserver5:~ # getfattr -R -d -e hex -m "trusted.afr." 
/mnt/gluster/brick?/storage | grep -v 0x000000000000000000000000 | grep 
-A1 -B1 trusted
getfattr: Removing leading '/' from absolute path names

# file: 
mnt/gluster/brick0/storage/images/1831/db88d55e-3282-c7c6-d1dd-ec41a665011f/hdd-images/8987
trusted.afr.storage0-client-0=0x000000000100000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1831/92f63f17-eb6c-8dba-2b9c-2e9cc52a8b2c/hdd-images/8786
trusted.afr.storage0-client-0=0x000000000100000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1831/6ae6c5eb-e6e2-4dfe-7bb3-75c622910f27/hdd-images/9113
trusted.afr.storage0-client-0=0x000000000100000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1828/4e5fd475-19b3-c9a7-1ad0-e4da528e6dbd/iso-images/11091
trusted.afr.storage0-client-0=0x000000000200000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1853/3df576b8-4206-e45a-33d7-433d56b700f0/iso-images/1957
trusted.afr.storage0-client-0=0x000000000200000000000000

# file: 
mnt/gluster/brick0/storage/images/1853/3df576b8-4206-e45a-33d7-433d56b700f0/iso-images/1960
trusted.afr.storage0-client-0=0x000000000200000000000000
--
# file: 
mnt/gluster/brick0/storage/images/2003/5e9b2bdc-a158-796f-6d81-60f39aee5137/hdd-images/9772
trusted.afr.storage0-client-0=0x000000000200000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1962/5c4cd738-bb56-d723-f001-0428e55ea81b/iso-images/8110
trusted.afr.storage0-client-0=0x000000000200000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1787/a190f24c-ed40-5642-4226-f00726dfc99f/iso-images/9837
trusted.afr.storage0-client-0=0x000000000200000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1787/ad8179fd-4f00-086c-955f-c2e469809e64/iso-images/2854
trusted.afr.storage0-client-0=0x000000000200000000000000

# file: 
mnt/gluster/brick0/storage/images/1787/ad8179fd-4f00-086c-955f-c2e469809e64/iso-images/10703
trusted.afr.storage0-client-0=0x000000000200000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1787/1782da17-059c-d159-373a-9ad9f5f9289f/iso-images/8519
trusted.afr.storage0-client-0=0x000000000200000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1787/aed64fba-2372-6f06-0690-be46136464a0/iso-images/10258
trusted.afr.storage0-client-0=0x000000000200000000000000

# file: 
mnt/gluster/brick0/storage/images/1787/aed64fba-2372-6f06-0690-be46136464a0/iso-images/10452
trusted.afr.storage0-client-0=0x000000000200000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1787/cd6089d7-c2cd-a5e1-2130-770fe028b5e3/iso-images/10511
trusted.afr.storage0-client-0=0x000000000200000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1834/02d04e16-40db-d244-aa4e-3e53cfaa2405/iso-images/504
trusted.afr.storage0-client-0=0x000000000200000000000000
--
# file: 
mnt/gluster/brick0/storage/images/1978/21527903-ca4e-4715-b40b-30c150f86d44/iso-images/9275
trusted.afr.storage0-client-0=0x000000000100000000000000

# file: 
mnt/gluster/brick0/storage/images/1978/21527903-ca4e-4715-b40b-30c150f86d44/iso-images/9511
trusted.afr.storage0-client-0=0x000000000100000000000000
--
# file: 
mnt/gluster/brick1/storage/images/1828/fc701d50-0b29-7827-89c0-77134ba96205/iso-images/9442
trusted.afr.storage0-client-2=0x000000000200000000000000
--
# file: 
mnt/gluster/brick1/storage/images/1878/875ed0c0-38b3-4552-7f1f-49a619996e5c/hdd-images/5758
trusted.afr.storage0-client-3=0x00000000ffffffff00000000
--
# file: 
mnt/gluster/brick1/storage/images/1787/ad8179fd-4f00-086c-955f-c2e469809e64/iso-images/2857
trusted.afr.storage0-client-2=0x000000000200000000000000

# file: 
mnt/gluster/brick1/storage/images/1787/ad8179fd-4f00-086c-955f-c2e469809e64/iso-images/10773
trusted.afr.storage0-client-2=0x000000000200000000000000

# file: 
mnt/gluster/brick1/storage/images/1787/ad8179fd-4f00-086c-955f-c2e469809e64/iso-images/10648
trusted.afr.storage0-client-2=0x000000000200000000000000
--
# file: 
mnt/gluster/brick1/storage/images/2003/8d7880ff-e7b2-3996-3fa3-ddb8022ca403/iso-images/9979
trusted.afr.storage0-client-2=0x000000000200000000000000
--
# file: 
mnt/gluster/brick1/storage/images/2003/116a4a8f-c8c2-6f70-1256-b29477d65e72/iso-images/10587
trusted.afr.storage0-client-2=0x000000000200000000000000
--
# file: 
mnt/gluster/brick1/storage/images/1956/ff4bbfd7-3b1a-00da-c901-c35cd967b600/iso-images/6815
trusted.afr.storage0-client-2=0x000000000200000000000000
--
# file: 
mnt/gluster/brick1/storage/images/1831/d56262c6-7f53-ec1b-baaf-15ceee9914ac/iso-images/9196
trusted.afr.storage0-client-2=0x000000000100000000000000
--
# file: 
mnt/gluster/brick1/storage/images/1959/cd55c5f3-9aa1-bfd9-99a0-01c13a7d8559/hdd-images/9324
trusted.afr.storage0-client-2=0x000000000200000000000000/



All other three server are happy, just pserver5 seems to be "stuck" ...

BTW; is there a way to relate the client number to a real machine? So, 
how do I check which machine is "client-2" or ""client-0"?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20110428/b20352aa/attachment.html>


More information about the Gluster-users mailing list